Qwen: Qwen3 Coder

Text input Text output
Author's Description

Qwen3-Coder-480B-A35B-Instruct is a Mixture-of-Experts (MoE) code generation model developed by the Qwen team. It is optimized for agentic coding tasks such as function calling, tool use, and long-context reasoning over repositories. The model features 480 billion total parameters, with 35 billion active per forward pass (8 out of 160 experts). Pricing for the Alibaba endpoints varies by context length. Once a request is greater than 128k input tokens, the higher pricing is used.

Key Specifications
Cost
$$$
Context
1M
Parameters
480B
Released
Jul 22, 2025
Speed
Ability
Reliability
Supported Parameters

This model supports the following parameters:

Presence Penalty Top P Tool Choice Temperature Seed Tools Structured Outputs Response Format Max Tokens
Features

This model supports the following features:

Tools Structured Outputs Response Format
Performance Summary

Qwen3 Coder, a Mixture-of-Experts (MoE) model optimized for agentic coding tasks, demonstrates competitive overall performance. Its speed ranks at the 50th percentile, indicating it performs comparably to many other models in terms of response times. Pricing is also competitive, sitting at the 49th percentile. A significant strength is its exceptional reliability, achieving a 100% success rate across all benchmarks, meaning it consistently provides usable responses without technical failures. In terms of specific benchmark performance, Qwen3 Coder shows strong capabilities in classification tasks, achieving 98.0% accuracy in Email Classification, placing it in the 62nd percentile for this category. While its cost for this task is competitive, its duration is somewhat higher. For Instruction Following, the model achieved 65.4% accuracy, ranking in the 74th percentile, suggesting a solid ability to interpret and execute complex instructions. However, the duration for this benchmark was notably long, placing it in the 33rd percentile, indicating a potential area for optimization in processing highly complex, multi-step instructions. Overall, Qwen3 Coder excels in reliability and classification accuracy, while offering competitive pricing and moderate speed. Its primary area for improvement lies in the speed of processing very long and complex instruction sets.

Model Pricing

Current Pricing

Feature Price (per 1M tokens)
Prompt $0.2
Completion $0.8

Price History

Available Endpoints
Provider Endpoint Name Context Length Pricing (Input) Pricing (Output)
Alibaba
Alibaba | qwen/qwen3-coder-480b-a35b-07-25 1M $0.2 / 1M tokens $0.8 / 1M tokens
Hyperbolic
Hyperbolic | qwen/qwen3-coder-480b-a35b-07-25 262K $2 / 1M tokens $2 / 1M tokens
Parasail
Parasail | qwen/qwen3-coder-480b-a35b-07-25 262K $0.39 / 1M tokens $1.6 / 1M tokens
Targon
Targon | qwen/qwen3-coder-480b-a35b-07-25 262K $1 / 1M tokens $2 / 1M tokens
Alibaba
Alibaba | qwen/qwen3-coder-480b-a35b-07-25 1M $0.2 / 1M tokens $0.8 / 1M tokens
Alibaba
Alibaba | qwen/qwen3-coder-480b-a35b-07-25 262K $1.5 / 1M tokens $7.5 / 1M tokens
DeepInfra
DeepInfra | qwen/qwen3-coder-480b-a35b-07-25 262K $0.2 / 1M tokens $0.8 / 1M tokens
Chutes
Chutes | qwen/qwen3-coder-480b-a35b-07-25 262K $0.2 / 1M tokens $0.8 / 1M tokens
Novita
Novita | qwen/qwen3-coder-480b-a35b-07-25 262K $0.2 / 1M tokens $0.8 / 1M tokens
Novita
Novita | qwen/qwen3-coder-480b-a35b-07-25 262K $0.64 / 1M tokens $2.5 / 1M tokens
Together
Together | qwen/qwen3-coder-480b-a35b-07-25 262K $2 / 1M tokens $2 / 1M tokens
DeepInfra
DeepInfra | qwen/qwen3-coder-480b-a35b-07-25 262K $0.3 / 1M tokens $1.2 / 1M tokens
Chutes
Chutes | qwen/qwen3-coder-480b-a35b-07-25 262K $0.2 / 1M tokens $0.8 / 1M tokens
GMICloud
GMICloud | qwen/qwen3-coder-480b-a35b-07-25 131K $1 / 1M tokens $2 / 1M tokens
BaseTen
BaseTen | qwen/qwen3-coder-480b-a35b-07-25 262K $0.38 / 1M tokens $1.53 / 1M tokens
Phala
Phala | qwen/qwen3-coder-480b-a35b-07-25 262K $0.9 / 1M tokens $1.5 / 1M tokens
Cerebras
Cerebras | qwen/qwen3-coder-480b-a35b-07-25 131K $0.2 / 1M tokens $0.8 / 1M tokens
Cerebras
Cerebras | qwen/qwen3-coder-480b-a35b-07-25 131K $2 / 1M tokens $2 / 1M tokens
AtlasCloud
AtlasCloud | qwen/qwen3-coder-480b-a35b-07-25 262K $0.7 / 1M tokens $2.5 / 1M tokens
Fireworks
Fireworks | qwen/qwen3-coder-480b-a35b-07-25 262K $0.45 / 1M tokens $1.8 / 1M tokens
Benchmark Results
Benchmark Category Reasoning Free Executions Accuracy Cost Duration
Other Models by qwen